In 375 women who accepted contraception at a clinic in northern Nigeria there were significant differences between IUD and pill acceptors. The pill acceptors were younger women of lower parity who tended to be less educated and mostly not engaged in any profession. The IUD clients were older women of high parity with better education and many had occupations outside the home. The most striking difference was in the religious affiliations of the two groups. The majority of Muslim women (72·9%) opted for the pill whereas the majority of Christian women (55·9%) chose the IUD.

In a 1982 survey of 250 men living in Khartoum, knowledge of, attitudes towards and use of contraception among Sudanese men were explored. Particular attention was given to attitudes towards male and female sterilization. Although most of the men expressed a desire to have large families, several findings suggest that they felt that they may have to be satisfied with a lower number for economic reasons. Three-fifths of the men with wives of reproductive age wished to use family planning services now but only about one-fifth were currently using an effective method. Almost 60% of the men were interested in learning more about female sterilization and about 50%, more about male sterilization. In general, attitudes towards family planning were more favourable than previously believed by providers of these services. Only 2·8% of these men had obtained services from a family planning clinic, suggesting that the demand currently exceeds the supply.

Data collected in Papua New Guinea between November 1979 and March 1980 show that the main reason for prolonged breast-feeding and sexual abstinence is to ensure the welfare of the mother and child (which is indirectly a means of child spacing). The mean period of breast-feeding for the rural mothers is 21·8 months and for the urban mothers is 20·7 months. The mean duration of sexual abstinence for the rural male respondents is 21·4 months and for the female is 20·2 months. The figures for their urban counterparts are 19·5 months and 16·6 months respectively.

In the area studied in the Machakos Project the total fertility estimate amounts to 7·46, which is somewhat lower than reported in the preceding article (van Ginneken et al., 1984). Evidence is provided that the low observed infant mortality rate of 49 per 1000 live births is plausible; this rate is in accordance with the relatively low and rapidly declining infant mortality rates found in Kenya. No accurate estimate on adult mortality could be obtained, probably due to serious overstatement of ages in the older age groups.

This paper analyses the impact of the length of birth intervals on mortality and growth of young children. An attempt has been made to improve the methodology of birth interval studies and the results of an analysis of longitudinal data from a rural area in Kenya with a relatively favourable level of health are presented. It is shown that children with short retrospective or short prospective birth intervals do not run a higher risk of mortality or growth retardation than children with longer intervals, neither during the perinatal period nor during the first 2 years of life. This suggests that the socioeconomic setting, in which a birth interval study is carried out, is of ultimate importance.

The conventional view that childlessness was associated with higher than average risk of marital breakdown was questioned by Chester (1972) and Gibson (1980) on the basis of analysis of divorce records. Data from the 1980 General Household Survey are used to show that childless couples, and those who have children early in marriage (and are thus more likely to have large families) are considerably more at risk of breakdown at all marriage durations than delayed two-child families. The reasons for the different conclusions of these three studies are assessed and it is concluded that standardization has many unattractive features for demographic analysis (especially in areas other than mortality), and that presentations based on relative or absolute risk of breakdown allow the key features to be assessed most satisfactorily.

Twenty-seven voluntarily childless couples who had made early decisions prior to marriage not to have children were compared to 23 couples childless through a series of successive postponements of childbearing. While the two groups differed little on background characteristics, husbands in early deciding relationships were less feminine-oriented and their wives more masculine-oriented than husbands and wives in the postponing group. Early deciding couples revealed higher levels of commitment to their choice, especially being more likely to seek an abortion for any pregnancy or to adopt out any child. A repertory grid measure revealed few differences in perceptions of others by either group of deliberately childless couples.

Following a suggestion that in Swansea students the downward trend of mean menarcheal age in Britain had ceased, data were collected on girls admitted to the University of Warwick over an 11-year period. There appears a slight increase in menarcheal age during the course of the study, similar to the upward trend seen in the latter part of the Swansea survey. This is not due to any change in the regional representation of the students, or in any of the other variables examined. Menarcheal age is not affected by father's socioeconomic status. It tends to be delayed in girls of larger families, while girls born later in a family of a given size have earlier menarche. Girls of more linear physique tend to have later menarcheal age.

Four hundred and fourteen adoptees were located in a population of 28,879 males born within a 4-year period. For all of these males, two social class ratings were obtained, based upon their occupations at age 25–28 and 35–38. A single rating was obtained for their fathers, based upon occupation at the time of birth of the population males. In the case of adoptees, this rating was obtained for both the biological and the adoptive fathers. The adoptees had, at both ages, an average social class not deviating from that of the population at large. Their biological fathers were, however, below average paternal social class and their adoptive fathers were above it. Positive correlations for social class were found between the adoptees, at both ages, and their biological and adoptive fathers. The social class of adoptees is less well predicted by that of their biological and adoptive fathers, even when these are taken jointly, than the social class of sons in the population is predicted by that of their fathers. Evidence from both the group means and the correlations suggests that the adoptees at age 35–38 came to resemble their biological fathers in social class more than they had done at age 25–28.

This study investigated factors influencing child survival to age 5 years for a rural farm sample in Iloilo Province, Philippines. Children from better quality housing were more likely to live to age 5 than were children from worse quality housing. The life expectancy at birth, indirectly estimated by the Brass method, was 63·4 years for children from worse quality housing and 69·9 years for children from better quality housing. The implications are explored.